Article pubs.acs.org/est
Evaluation of the Genotoxic and Physiological Effects of Decabromodiphenyl Ether (BDE-209) and Dechlorane Plus (DP) Flame Retardants in Marine Mussels (Mytilus galloprovincialis) Enrique Barón,† Awantha Dissanayake,‡ Judit Vilà-Cano,† Charlotte Crowther,‡ James W. Readman,‡,§,# Awadhesh N. Jha,‡ Ethel Eljarrat,*,† and Damià Barceló†,⊥ †
Institute of Environmental Assessment and Water Research (IDAEA), Spanish Council for Scientific Research (CSIC), Jordi Girona 18-26, 08034 Barcelona, Spain ‡ School of Biological Sciences, Plymouth University, Drake Circus, Plymouth PL4 8AA, United Kingdom § School of Geography, Earth & Environmental Sciences, Plymouth University, Drake Circus, Plymouth PL4 8AA, United Kingdom # Plymouth Marine Laboratory, Prospect Place, The Hoe, Plymouth PL1 3DH, United Kingdom ⊥ Catalan Institute for Water Research (ICRA), H2O Building, Scientific and Technological Park of the University of Girona, Emili Grahit 101, 17003 Girona, Spain S Supporting Information *
ABSTRACT: Dechlorane Plus (DP) is a proposed alternative to the legacy flame retardant decabromodiphenyl ether (BDE209), a major component of Deca-BDE formulations. In contrast to BDE-209, toxicity data for DP are scarce and often focused on mice. Validated dietary in vivo exposure of the marine bivalve (Mytilus galloprovincialis) to both flame retardants did not induce effects at the physiological level (algal clearance rate), but induced DNA damage, as determined by the comet assay, at all concentrations tested. Micronuclei formation was induced by both DP and BDE-209 at the highest exposure concentrations (100 and 200 μg/L, respectively, at 18% above controls). DP caused effects similar to those by BDE-209 but at lower exposure concentrations (5.6, 56, and 100 μg/L for DP and 56, 100, and 200 μg/L for BDE-209). Moreover, bioaccumulation of DP was shown to be concentration dependent, in contrast to BDE-209. The results described suggest that DP poses a greater genotoxic potential than BDE-209.
■
INTRODUCTION Polybrominated diphenyl ethers (PBDEs) were one of the most used halogenated flame retardants (HFRs) worldwide, available as three main commercial mixtures: Penta-BDE, OctaBDE, and Deca-BDE.1 However, this situation has changed due to recent restrictions over PBDEs. Within the European Union (EU), Penta- and Octa-BDE mixtures were banned in 2004, whereas Deca-BDE mixture was banned in 2008.2 PBDEs have been found in a wide range of environmental matrices such as sediment, water, fish or cetaceans, and also in humans.3−8 Nonetheless, the environmental behavior and effects of BDE209 have been studied to a lesser extent than those of lower brominated PBDEs. This might be due to the limitations in the analytical methodologies for the analysis of this compound in the past, due to its high log Kow and molecular weight.9 Despite the bioaccumulation potential being lower than those of other low brominated PBDEs such as BDE-47, BDE-209 has been found in different vertebrate and invertebrate species worldwide.2,10,11 In fact, BDE-209 was the main PBDE found in species with terrestrial diet2,10,12 and also in mussels.11,13 BDE© 2016 American Chemical Society
209 has shown thyroid and endocrine disruption properties,14,15 and it could affect the liver of fish and mice.15,16 Most of the studies are focused in vertebrate models such as fish or mice;17 thus, studies in invertebrates, for instance mussels, are scarce. Dechlorane Plus (DP) was selected as an alternative to Mirex when it was banned as a flame retardant (FR), and currently it has been proposed as an alternative to the Deca-BDE mixture. It is considered a novel HFR and is still barely regulated.18−20 Similar to BDE-209, DP has been found in a wide range of biological matrices such as fish, mussels, or cetaceans and also in humans, showing its bioaccumulation capacity.18−21 Toxicity data for DP are still very scarce.22,23 In fish, DP affected protein responses in the liver and induced apoptosis,24 whereas it Received: Revised: Accepted: Published: 2700
November 25, 2015 January 24, 2016 February 1, 2016 February 1, 2016 DOI: 10.1021/acs.est.5b05814 Environ. Sci. Technol. 2016, 50, 2700−2708
Article
Environmental Science & Technology showed genotoxical potential in bacteria 22 as well as histopathological changes in mice liver.25 Mussels have proven to be a good tool to evaluate the environmental behavior of organic pollutants.26 Furthermore, effects of organic pollutants in mussels have been correlated with effects of the same pollutants in humans,27 which shows that these contaminants can affect the whole food chain. Thus, the study of the effects of FRs in mussels could provide useful information concerning the potential for effects of these contaminants in other biota and ecosystems. Consequently, the aim of this study was to evaluate the genotoxic and physiological effects of one classical FR (BDE-209, which represents about 98% of the Deca-BDE commercial mixture) and one alternative FR commercial mixture (DP) in Mytilus galloprovincialis through an in vivo exposure via the dietary pathway. To our knowledge, this is the first time that the toxicity of DP has been evaluated in this way. M. galloprovincialis is predominately native to the Mediterranean coast and the Black and Adriatic Seas; however, it has established itself as a global invader. This species has highly conserved gene sequences shared by higher organisms including humans as described by us in a previous study.28 Effects reported in this model invertebrate would therefore have significance for higher level impacts in coastal environments and could be translated to other species.
model by which water was changed every day and mussels fed daily. B(a)P was chosen as a model organic contaminant as it is relatively insoluble in water (Log Kow = 6.04), known to cause genetic damage, and a priority pollutant.28 After the dietary pathway proved to be a valid exposure route, mussels were exposed to three different concentrations of BDE-209 (56, 100, and 200 μg L−1) and DP (5.6, 56, and 100 μg L−1), following the procedure described above; that is, individual mussels (n = 7 for each concentration treatment) were placed in 2 L beakers containing 1.8 L of filtered seawater. Alga mortality was evaluated before the first exposure by exposing an aliquot of the algae to seawater, acetone, acetone + DP, and acetone + BDE-209. No changes in cell size were observed. Concentrations of DP found in the environment and specifically in mussels are considerably lower than concentrations found for BDE-209. Thus, exposure concentrations of DP were settled in lower scale, although two common concentrations were maintained for comparisons. Environmental concentrations of these contaminants in mussels greatly depend on the sampling area. For instance, DP has been found at levels up to 190 ng/g lipid weight (lw) in an industrial area of China, but concentrations in rural areas were considerably lower, 4.1 ng/g lw.29,30 These concentrations are of the same magnitude (parts per billion, ppb) as concentrations selected in this study. With regard to BDE-209, concentrations reported worldwide vary substantially. BDE-209 has been reported at concentrations up to 812 ng/g dw in sediments,31 and it is detected consistently in wild mussels.32 Hence, it is present at important levels in the environment and bioconcentrates. The selected exposure concentrations are higher than some reported globally, but are on the same order as others. A B(a)P exposure at 100 μg L−1 as positive in vivo control together with a negative control (acetone, 0.05%, v/v final volume) were also performed (n = 7 per treatment). H2O2 was also used as positive control for in vitro validation study for comparison (1 mM and 30 min of exposure time). In both experiments, after the 6 days of exposure, mussel hemolymph was extracted from the posterior adductor muscle using an ice-chilled 1 mL syringe and 21G needle and transferred into individual Eppendorf tubes held on ice, following the protocol described by Brown et al.33 Water Quality. Water quality (temperature, salinity, dissolved oxygen, and pH) was measured every day for each beaker, and three water samples of each treatment were taken immediately after dosing and prior to water change (i.e., after 23 h of exposure). Water temperature during the exposure period was 16.0 ± 0.5 °C, salinity was 36.3 ± 0.2‰, dissolved oxygen was 7.93 ± 0.2 mg/L, and pH was 7.92 ± 0.08. No intra- or interday variations among treatments were observed (ANOVA and post hoc Tukey’s test), and these values were considered optimal for the exposures. Clearance Rate (CR). CR was determined prior to hemolymph collection as described previously.34 Mussels were placed in separate 400 mL beakers containing 350 mL of seawater (filtered to 0.8 μm) and a stirring bar. They were allowed to acclimatize at 15 °C for 15 min. I. galbana was added in a concentration of 10,000 cells/mL, including several procedural blanks (beaker plus 300 mL of seawater). Aliquots of 20 mL were removed immediately after the addition and after 10, 20, and 30 min. These aliquots were analyzed on a Beckman Coulter Particle Size and Count Analyzer set to count particles between 4 and 10 μm. Three separate counts per
■
MATERIALS AND METHODS Sample Collection. M. galloprovincialis samples (5−6 cm length) were collected during the last week of July 2014 from Trebarwith Strand (North Cornwall, UK), one of the most pristine sites in the United Kingdom, and were immediately transported to the laboratory, rinsed with seawater, and acclimatized in an aerated tank with 50 L of filtered seawater (0.8 μm), where they were maintained at 15 ± 1 °C with a photoperiod of 12 h of light/12 h of darkness for 10 days and fed every 2 days with Isochrysis galbana (Liquifry, Interpet, Dorking, UK). Stocking density was 3 mussels per liter. Water was changed 2−3 h after feeding. Any spawning animals were removed from the holding conditions, and no animals spawned during the experiments. Chemicals and Reagents. Triton X-100, sodium chloride, normal melting point agarose (NMPA), low melting point agarose (LMPA), and N-laurylsarcosine were purchased from Sigma-Aldrich (UK). BFR-PAR solution, containing BDE-28, BDE-47, BDE-99, BDE-100, BDE-154, BDE-183, and BDE209, and syn- and anti-DP were purchased from Wellington Laboratories (Guelph, ON, Canada), as well as the internal standard 13C-BDE-209. 13C-syn-DP, used also as internal standard, was obtained from Cambridge Isotope Laboratories (Andover, MA, USA). Experiment. To assess whether the feeding route was a valid exposure pathway for filter-feeding organisms when exposed to high Log Kow organic contaminants, a preliminary experiment using benzo(a)pyrene (B(a)P) was performed. The genotoxic potential of this polycyclic aromatic hydrocarbon (PAH) is well-known, and it is often used as a genotoxic model.28 Individual mussels were placed in 2 L beakers containing 1.8 L of filtered seawater and exposed to B(a)P at either 100 or 200 μg L−1 for 6 days, each concentration dosed either by spiking algae I. galbana or directly into the aqueous media (n = 6 per concentration treatment, including a solvent carrier (acetone, 0.05% v/v) control with only acetone). Both exposure pathways were conducted following a semi-static 2701
DOI: 10.1021/acs.est.5b05814 Environ. Sci. Technol. 2016, 50, 2700−2708
Article
Environmental Science & Technology Table 1. Recoveries, RSD, MDLs and MQLs of BDE-209 and DP in Water and Mussel water
mussel
compound
recovery (%)
RSD (%)
MDL (ng/mL)
MQL (ng/mL)
recovery (%)
RSD (%)
MDL (pg/g lw)
MQL (pg/g lw)
BDE-209 syn-DP anti-DP
75 67 73
11 8 12
0.40 0.60 0.10
1.30 2.00 0.30
68 85 88
6 7 5
200 5.50 4.30
330 18.3 14.3
hemocytes with induced MN were carefully distinguished from hemocytes with nuclear buds; the latter were not counted.42 Chemical Analysis. With regard to water and algae analysis, the methodology described by Di et al. was adopted.28 Hexane (1 mL) was added to 9 mL of the exposure water samples, and internal standards (13C-BDE-209 and 13C-syn-DP) were added. Samples were manually shaken and then centrifuged at 3500 rpm for 10 min. The aqueous phase was discarded, and the organic phase was evaporated to dryness and then reconstituted to a final volume of 500 μL with toluene. Mussel samples were extracted using a previously described methodology.43,44 Briefly, samples were spiked with 100 ng of 13 C-BDE-209 and 13C-syn-DP and kept overnight to equilibrate prior to extraction by pressurized liquid extraction (PLE). Afterward, lipid content was determined gravimetrically and redissolved in hexane prior to acid treatment (H2SO4(c)). A solid phase extraction (SPE) using Al−N cartridges (Biotage, 5 g and 20 mL) was performed to complete the cleanup, and resulting extracts were concentrated to a final volume of 40 μL. Instrumental analysis was carried out using gas chromatography coupled to mass spectrometry working in negative chemical ionization mode (GC-NCI-MS) using an Agilent Technologies 7890A GC system coupled to 5890A GC-MS single quadrupole, following previously optimized protocols.45,46 BDE-209 was analyzed using NH3 as reagent gas, whereas DP was analyzed using CH4 as reagent gas. Selected ion monitoring (SIM) was used to enhance sensitivity. Two ions were monitored for each compound: the most intense was used for quantification and the second for confirmation. Ions monitored were m/z 487 and 489 for BDE-209 (497 and 499 for 13C-BDE-209) and m/z 654 and 656 for DP (664 and 666 for 13C-syn-DP). Recoveries, method detection limits (MDLs), and method quantification limits (MQLs) are shown in Table 1. Recoveries were determined by spiking 1 g of individual mussel samples with 10 ng of syn- and anti-DP and 50 ng of BDE-209. Five replicates were made, together with three blank samples. MDLs and MQLs were determined as the concentrations that gave signal-to-noise ratios (S/N) of 3 and 10, respectively. Statistical Analysis. Data were tested for normality and homogeneity of variances using the Shapiro−Wilks test of normality and an F test. Statistical significance between different treatments was determined using analysis of variance (ANOVA), post hoc Tukey’s test, and t test; a p value ≤0.05 was used to determine significant differences. Statistical analyses were conducted using the open-source statistical programming language R v. 3.1.1 (http://cran.r-project.org).
mussel were made. CR was calculated using the equation CR = V(log C1 − log C2)/t, where V is the volume of water, C1 and C2 are the cell concentrations at the beginning and end of each increment, respectively, and t corresponds to the time interval.35 Comet Assay. Determination of DNA strand breaks using hemocytes was evaluated following a previously optimized protocol.36,37 Slides were precoated with NMPA and kept overnight at 20 °C to dry. One hundred and fifty microliters of hemolymph was centrifuged at ∼350g at 4 °C for 2 min and then mixed with 150 μL of molten LMPA. Two separate drops of 75 μL were placed on the slide and immediately covered with a coverslip. Prior to performance of the comet assay, cell viability was determined using Eosin Y staining;38 viability was deemed >95%. Slides were kept at 4 °C and in the dark for 1 h to allow the gel to solidify. In the case of the H2O2 in vitro positive control, after 1 h, 1 mL of H2O2 (1 mM) was added dropwise and incubated at 4 °C for 30 min. Slides were incubated in lysis solution for 1 h and in the dark at 4 °C, placed in the electrophoresis chamber, filled with electrophoresis buffer, and incubated for 20 min to unwind. Afterward, the chamber was turned on (25 V, 400 mA) and electrophoresis performed for 20 min. Subsequently, slides were neutralized with cold neutralization buffer. All of the steps in the electrophoresis procedure were performed at 4 °C and in the dark. Slides were stained with ethidium bromide (20 μL of a 20 μg/mL solution in each drop) and scored under an epifluorescense microscope (Leica, DMR) using the Komet 5 software (Kinetic Imaging, Nottingham, UK). Fifty cells in each drop, thus a total of 100 cells per slide, were scored, and % tail DNA was used for the evaluation of DNA strand breaks, because it has been validated through interlaboratory comparisons.39,40 In total, 7 slides per treatment for a total of 63 slides (3 DP treatments, 3 BDE-209 treatments, 1 B(a)P treatment, 1 negative control, and 1 H2O2 treatment) were analyzed. Abnormal comets were excluded from the scoring following the criteria proposed previously.41 In short, cells outside the gel, double cells, or comets in contact with other comets were not scored, and only comets with one round head on the backmost side in the direction of the analysis were scored. Micronuclei (MN) Assay. Induction of MN in hemocytes was evaluated as described by Jha et al.37 Slides were previously coated with 10% poly-L-lysine solution and dried overnight. Two hundred microliters of hemolymph was spread gently onto the slide and left at 15 °C for 30 min and then fixed with MeOH for 15 min. Afterward, slides were stained using Giemsa stain (5%, v/v) for 20 min; excess stain was removed with MilliQ water, and once the slides were air-dried, a coverslip was mounted using DPX. Slides were scored randomly under the microscope for the induction of MN. Approximately 1000 cells from each slide were scored following the criteria described in previous works.37 In total, 63 slides (7 per treatment) were analyzed. Only agranular cells were scored, and apoptotic and necrotic hemocytes were excluded from the analysis. Moreover,
■
RESULTS AND DISCUSSION In Vitro Validation of Comet Assay Using H2O2. Various concentrations (0.2, 0.5, and 1 mM) and time points (10 and 30 min) were explored to validate H2O2 induced DNA damage. Results showed that DNA damage due to H2O2 exposure in vitro is time-dependent with significantly more DNA damage apparent at the longer time point (ANOVA, p < 2702
DOI: 10.1021/acs.est.5b05814 Environ. Sci. Technol. 2016, 50, 2700−2708
Article
Environmental Science & Technology
are in agreement with previous studies.28,47,48 For instance, Di et al. report 60% damage following a 12 day in vivo exposure of Mytilus edulis.28 However, DNA strand breaks in control mussels were 30% and, thus, DNA relative damage induced by B(a)P was up to 30%, similar to our reported values. Clearance Rate. It has been previously demonstrated that the CR in mussels can be affected by several chemical contaminants.34 In this experiment, the CR ranged from 0.49 to 0.90 L/h in the first time increment (10 min) for both BDE209 and DP, whereas it was 0.46 and 0.69 L/h for B(a)P and control treatments, respectively. No statistical differences were found among the treatments, although all of them were significantly different from the seawater control (ANOVA, F2.78 = 3.196, p < 0.05). The same scenario occurred in the second time increment (20 min), where the CR value increased to 1.64−2.16 for BDE-209 and DP, to 1.29 for B(a)P, and to 1.57 in controls. Even though values for BDE-209 (100 μg/L) and DP (56 μg/L) increased more rapidly than other treatments, differences were not significant with any treatment with FR. Finally, after 30 min, CR reached values ranging from 1.98 to 2.92 L/h for both BDE-209 and DP, 1.77 L/h for B(a)P, and 2.09 for control mussels. Again, even if BDE-209 (100 μg/L) and DP (56 μg/L) showed higher values than the other treatments, these differences were not significant (Figure 2). Thus, we can summarize that mussels are not significantly affected by these FRs at a physiological level, at least with the end point chosen in this study. This fact was described for B(a) P in a similar experiment28 and suggests that mussels can take up these types of compounds without showing significant physiological changes.38 Comet Assay. In all cases, DNA strand breaks observed were significantly higher than the negative control (ANOVA and Tukey’s test, p < 0.001) (Figure 3). In vivo positive control, B(a)P, caused an effect of 35 ± 6% (mean % tail ± SD), whereas the in vitro positive control, H2O2, resulted in 56 ± 10%. Damage induced by BDE-209 was 13 ± 3, 21 ± 3, and 21 ± 6% for 56, 100, and 200 μg/L exposure concentrations, respectively. Damage induced by DP was 13 ± 4, 23 ± 3, and
0.001). On the basis of these data, both in the pathway validation and in the main experiment in vitro controls were concurrently performed using a concentration of 1 mM and 30 min of exposure time. Dietary Pathway Validation. DNA damage was observed in all B(a)P-exposed mussels, irrespective of exposure route (diet or aqueous), and was significantly different from control mussels (ANOVA, p < 0.001) (Figure 1). The solvent control
Figure 1. DNA damage (mean ± SD; n = 6 per treatment) in benzo(a)pyrene-exposed mussels. Treatments with the same letter are not significantly different; when significant differences occur between treatments, p < 0.001.
exhibited a small amount of DNA damage (